QRNNs: "upto 16 times faster than LSTMs". Recently used in Baidu's Deep Voice https://metamind.io/research/new-neural-network-building-block-allows-faster-and-more-accurate-text-understanding …
What are some recent examples of successes of 1D convnets on tasks where RNNs were previously dominant, besides WaveNet and ByteNet?
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Quasi -rnns?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
We used it really well on mocap data where previous rnn was not working well http://theorangeduck.com/page/deep-learning-framework-character-motion-synthesis-and-editing …
-
Very cool work. And thanks a bunch for making that video!
End of conversation
New conversation -
-
-
Parsey McParseface comes to mind but it's feedforward https://arxiv.org/abs/1603.06042
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.